2026 CF Workshop: Call for breakout and hackathon sessions

[sadielbartholomew]

Dear CF Community,

We are happy to invite ideas for breakout sessions - either discussion-based or hackathon (practical/coding) sessions - for this year's annual CF Workshop, which will be held as a hybrid event 21-24 September 2026 based (for in-person attendees) at the ECMWCF site in Bonn, Germany (see #472).

We'd like to hear from anyone who wishes to either:

• offer to host a breakout on a specific topic themselves, or;
• suggest a breakout they would like to see as hosted by other(s) - please make a nomination if you have anyone in mind.

Breakouts may focus on topics already registered via the Issue Trackers on the 'cf-convention' repositories (notably 'discuss' and 'cf-conventions') or the Discussions forum, or new ones as yet unrecorded.

Inspiration might be found at the CF meetings home page, where you can see what talks were held at previous meetings. The 2026 Workshop agenda will be published on the CF website when the content of the workshop has been finalised.

Note that we are also inviting proposals for talks i.e. presentations (in #475), so consider whether you would like to propose a breakout session or a talk depending on what format is best: either a one-to-many presentation (talk) or a many-to-many discussion or practical (perhaps hackathon-style) session between a small group (breakout session).

We hope that the outcome of a breakout session will be one or more of:

• progress on some new, neglected, or stymied issue or opportunity of importance to CF (including admin-aligned matters such as repository housekeeping and information management / devops);
• recommendation on whether to pursue any CF proposals relevant to the breakout;
• more future contributors to CF.

Proposals can be short and informal. One or two paragraphs will suffice. Please make your proposal as a comment on the discussion thread for this announcement.

Contributions from those who are new to CF and/or to the CF Workshops are very welcome!

Sincerely,
Sadie, on behalf of the CF Workshop Organising Committee

[cofinoa]

Thanks @sadielbartholomew.

Perhaps it would help to look back at breakout, discussion, and hackathon themes from previous CF Workshops, going back to 2016. These are not proposed agenda items, but they may help the community identify topics that could be continued, revived, or developed further in 2026.

The list below is intended only as a rough retrospective. Some topics may have appeared as formal breakout or hackathon sessions, while others may have been discussed in talks, plenary sessions, workshop notes, or related agenda items. Corrections and additions would be very welcome.

Possible recurring themes include:

• Standard names and vocabularies: broadening standard names to other geoscience domains, standard names governance, precipitation-related standard names, CF standard names and ontologies, mapping standard names to I-ADOPT, vocabularies, and visualisations for standard names data (2016, 2017, 2018, 2020, 2021, 2022, 2023, 2024, 2025).

• Grids, geometries, and spatial representation: UGRID and unstructured grids, simple geometries / data footprints, DSG and timeseries, satellite swath data, CF-Radial, HEALPix grids, grid mappings, CRS/WKT, OGC CRS harmonisation, and mesh or boundary variables (2016, 2017, 2018, 2019, 2020, 2024, 2025).

• Data models, formats, and interoperability: CF and the netCDF enhanced data model, groups, strings and character encodings, netCDF-4 features, external variables, netCDF/Zarr, ncZarr, NetCDF-Schema, GRIB/CF mapping, WMO GRIB2 metadata, file formats / data model discussions, and dataset aggregation (2016, 2017, 2018, 2019, 2021, 2022, 2023, 2024, 2025).

• Governance, process, and repository workflow: CF governance and process, GitHub migration and workflow, moderation of proposals, requirements keywords such as MUST/SHOULD/MAY, BCP-14, GitHub processes and automation, release/process questions, roadmap / white paper development, and website / conventions housekeeping (2016, 2017, 2018, 2019, 2020, 2022, 2023, 2024, 2025).

• Metadata extensions and provenance: linked data, semantic technologies, metadata graphs, attribute namespaces / URI prefixes, external standards, discovery metadata, provenance metadata, metadata for statistical processing, metadata pipelines, trusted provenance, and metadata for new technologies (2016, 2017, 2018, 2020, 2021, 2024, 2025).

• Compression, uncertainty, and derived products: uncertainty metadata, climate indices, derived statistics, cell methods, bounds, lossy compression metadata, compression precision, and coordinate subsampling (2016, 2018, 2019, 2020, 2021, 2022, 2023, 2024).

• Tools, examples, validation, and training: CF checkers, reference implementations of the CF data model, CF tools, example datasets, compliance-related material, schema-based validation, training resources, and practical test material (2020, 2021, 2023, 2024, 2025).

• Persistent identifiers and documentation outputs: DOIs for the CF Conventions, dataset DOIs, DOI attributes, bibliography/PDF improvements, and more persistent workshop outputs (2018, 2019, 2021, 2022, 2023, 2024, 2025).

• AI / machine-learning related work: CF use cases for machine-learning training datasets, CF-NetCDF file headers for ML training, LLM-related approaches to CF metadata, and LLM-assisted standard name generation (2021, 2024, 2025).

This kind of retrospective list could provide a concrete starting point: people could say which previous topics still need work, which ones are now resolved, and which ones they would be willing to help host as a 2026 breakout or hackathon.

[TomLav]

This is useful, thank you. I want to mention that I have recently been involved in several projects (incl. for ESA) in which satellite swath data file formats were discussed. In general, my impression is that the CF convention itself is not a bottleneck (it has many / all of the necessary concepts), but there are no shared good practices / templates for how to apply the convention when creating new satellite swath data formats. As a result, there is a zoo of satellite swath file formats out there. There have been some attempts to constrain this in the past (e.g. cf-convention/cf-conventions#269) but they stalled or were abandoned. Maybe because it is not possible to make general rules. I could be interested in discussing this topic with other interested people/parties.